Luce di sincrotrone: generazione e...

VIII Scuola Nazionale di Luce di Sincrotrone - Frascati 2005

Luce di sincrotrone: generazione e proprietà

Giorgio MargaritondoVice-président pour les affaires académiques

Ecole Polytechnnique Fédérale de Lausanne (EPFL)


Programma:• Come costruire un’ottima sorgente di

raggi x utilizzando la relatività di Einstein• Alcuni esempi di applicazioni • Raggi x coerenti: una rivoluzione nella

radiologia• Dai sincrotroni ai laser a elettroni liberi e

al loro uso• Il futuro: nuove sorgenti, laser a raggi x

SASE


From ancient fires to synchrotrons and FEL’s, the same problems:

A fire is not very effective in "illuminating" a specific target: its emitted power is spread in all directions

A torchlight is much more effective: it is a small-size source with emission concentrated within a narrow angular spread -- it is a "bright" source

Likewise, we would like to use “bright” sources for x-rays (and ultraviolet light)


Why x-rays and ultraviolet?

Wave-length

(Å)

0.1

10

1000

Photonenergy(eV)

10000

1000

100

10

Coreelectrons

Valenceelectrons

Chemical bond

lengths

Molecules

Proteins


The “brightness” of a light source:

Flux, FAngular divergence, Ω

Source area, S

FS x Ω

Brightness = constant x _________


A real synchrotron facility: Swiss Light Source (SLS)


Objective: building a very bright x-ray source.Solution: relativity!!

Undulator Emitted x-rays

Circulating electrons (speed ≈c)

Ring under vacuum

• The undulator (periodic magnet array) period determines the emitted wavelength. This period is shortened by the relativistic “Lorentz contraction” giving x-ray wavelengths

• The emitted x-rays are “projected ahead” by the motion of their sources (the electrons), and therefore collimated. Relativity enhances the effect


Heat flux (watt/mm2)

100

10

1

Surface of the sun

Interior of rocket nozzle

Nuclear reactor core

Swiss Light Source (SLS)

ALS, Elettra, SRRC, PAL


The historical growth in brightness/brilliance(units: photons/mm2/s/mrad2, 0.1% bandwidth)

1021

1015

109

1900 1950 2000

Wigglers

Bendingmagnets

SLS

Rotatinganode

SLS (Swiss Light

Source)


Objective: building a very bright x-ray sourceDetails of the solution:

Undulator (periodic B-field,

period L ≈ centimeters

electron In the electron reference frame:• Periodic B-field → periodic B & E-fields

moving at speed ≈c, similar to electromagnetic wave

• Lorentz contraction: L → L/γ• Undulation of electron trajectory →

emission of waves with wavelength L/γ

Speed ≈ c

In the laboratory frame:• Doppler effect → wavelength further reduced by a factor

of ≈2γ, changing from L/γ to L/2γ2

Overall: L → L/2γ2

Centimeters → 0.1-1,000 Å (x-rays, UV)


What causes the high brightness?

• Free electrons can emit more light than bound electrons ⇒ high flux• The electron beam control is very sophisticated: small transverse

beam cross section ⇒ small synchrotron source size• Relativity collimates the emitted synchrotron radiation:

emitted x-ray

θ

Electron reference

framecx = dx/dt

cy = dy/dt

tn(θ) ≈ (cy/cx)

Laboratory frame

cx’ = dx’/dt’

cy’ = dy’/dt’electron velocity

Lorentz transformation: γ-factor for x’ and t’ but not for y’ ⇒ tn(θ) reduced by a factor ≈1/γ


3 types of sources:

detectorcontinuously illuminated

1. Undulators: smallundulations

detector

time

longsignal pulse

frequency

hν/∆hν≈ N

narrow hν-band


3 types of sources:

2. Bending magnets:

shortsignal pulse

broadhν-band

time frequency


3 types of sources:

3. Wigglers: largeundulations

broadhν-band

frequency

Series of short

pulses

time


3 types of sources - summary:

Undulators:

time frequency

More intensity than bending

magnetsWigglers:

frequencytime

Bending magnets:

time frequency


Undulator emission spectrum:

L = period

Central wavelength: L/2γ2

First correction: out of axis, the Doppler factor is not 2γ2 but changes with θ’Central wavelength: (L/2γ2)/(1+ 2γ2θ’2)

θ’

Second correction: higher B-field means stronger undulations and less on-axis electron speed. This changes γ so that:Central wavelength: (L/2γ2)/(1+ aB2)


Bending magnet emission spectrum:The (relativistic) rotation frequency of the electron determines the (Doppler-shifted) central wavelength: λo = (1/2γ2)(2πcmo/e)(1/B)

The “sweep time” δt of the emitted light cone determines the frequency spread δν and the wavelength bandwidth:∆λ / λo = 1

A peak centered at λcwith width ∆λ: is this really the well-known synchrotron spectrum?YES -- see the log-log plot:

λλ0

∆λ

log(λ)

λo

log(

emis

sion

)


Synchrotron light polarization:Electron in a storage ring:

TOP VIEW

TIL TED VIEW

SIDE VIEW

Polarization:Linear in the

plane of the ring, elliptical out of

the plane

Special (elliptical) wigglers and undulators can provide ellipticaly polarized light with high intensity


synchrotronradiation

atom ormolecule

scattered photons, fluorescence

small-angle scattering

fluorescence spectroscopy

photoelectrons, Auger electrons photoelectron/Auger

spectroscopy

transmitted photons

absorption spectroscopy

EXAFS

molecularfragments

fragmentation spectroscopy

solidscattered photons scattering

photoelectrons, Auger electronsphotoelectron/Auger spectroscopy

transmitted photons

absorption spectroscopy

EXAFS

fluorescence spectroscopyfluorescence

diffracted photons X-graphy

Atoms & molecules desorption spectroscopy

Synchrotron x-rays:Many different interactions

↓Many different applications


European Integrating Initiative on Synchrotron Radiation and Free Electron Laser Science


Synchrotron Facilitiesin the World (2005):

TOTAL: 72 in 24 Countries62 Operating10 under construction


Historical GrowthWorldwide ISI data 1968-2006, Keyword: “synchrotron”

0

5000

10000

15000

20000

25000

30000

1968 1978 1988 1998 2008


… and, for a broader picture:

Hits from a Google search:“Synchrotron” 5,540,000“Free electron laser” 3,020,000“Cyclotron” 1,850,000“LINAC” 1,860,000“Hadron collider” 730,000

“Protein crystallography” 1,890,000“Synchrotron photoemission” 176’000

“Viagra” 9’640’000“Pamela Anderson” 14,200’000


Photoelectron spectroscopy: basic ideas

Formation of chemical bonds:atom solid

elec

tron

ener

gyThe photon absorption

increases the electron energy by hν before ejection of the

electron from the solid

elec

tron

ener

gy

hνPhoton(hν)

Photoelectron

Photoelectric effect:


Photoelectron spectroscopy ofhigh-temperature superconductivity:

elec

tron

s

energy

normal state

super-conducting state

The limited energy resolution of conventional photoemission makes it impossible to observe the phenomenon

-0.2 -0.1 0Energy (eV)

High-resolution spectra taken with

ultrabright synchrotron

radiation


Superconducting gap spectroscopy:

Different gaps in different directions (d-wave superconductivity) (Kelly, Onellion et al.


From spectroscopy to spectromicroscopy:

Spectroscopy (energy and momentum

resolution)

Microscopy (spatial resolution)

Chemical information

Spectromicroscopy


The two modes of photoemission spectromicroscopy:

SCANNING ELECTRON IMAGING

X-rays

X-ray lenssample

e

x-y scanning stage

Electron analyzer

X-rays

Electron optics

e

sample


The ESCA Microscopy Beamline at ELETTRA


Photoelectron spectromicroscopy (on untreated specimens) beats optical

microscopy + staining in revealing cell nuclei(B. Gilbert , M. Neumann , S. Steen , D. Gabel , R. Andres, P. Perfetti, G.

Margaritondo and Gelsomina De Stasio)

The distribution of nuclei in human glioblastoma tissue, revealed (left) bystaining for optical microscopy and (right) by a MEPHISTO phosphorusmap on ashed tissue (phosphorus is shown dark). The MEPHISTO sectionon gold had no treatment other than ashing. The imaged areas are in adjacenttissue sections, so the exact pattern of nuclei distributions is not identical.

20µm


Coherence: “the property that enables a wave to produce visible diffraction and interference

effects”

fluorescent screen

screen with pinhole

θsource

(∆λ)

ξ

Example:

The diffraction pattern may or may not be visible on the fluorescent screen depending on the source size ξ, on its angular divergence θ and on its wavelength bandwidth ∆λ


Longitudinal (time) coherence:

source (∆λ)

• Condition to see the pattern: ∆λ/λ < 1• Parameter characterizing the longitudinal coherence:

“coherence length”: Lc = λ2/∆λ• Condition of longitudinal coherence: Lc > λ


Lateral (space) coherence — analyzed with a source formed by two point sources:

• Two point sources produce overlapping patterns: diffraction effects are no longer visible.

• However, if the two source are close to each other an overall diffraction pattern may still be visible: the condition is to have a large “coherent power” (2λ/ξθ)2


Coherence — summary:• Large coherence length Lc = λ2/∆λ• Large coherent power (2λ/ξθ)2

•Both difficult to achieve for small wavelengths (x-rays)

•The conditions for large coherent power are equivalent to the geometricconditions for high brightness


Some Problems in Conventional Radiology:

Low-intensity, divergent beam

Low absorption Limited contrast,

may require a high x-

ray dose


Light-matter Interactions:

Absorption -- described by the absorption coefficient α

Refraction (and diffraction/interference) --described by the refractive index n

For over one century, radiology was based on absorption: why not on refraction /diffraction?


Conventional radiology

Refractive-index radiology


“Refraction” x-ray imaging:

Edge between regions with different n-values

detector Detectedintensity

Idealized edge image

Real example (leaf)


“Refraction” x-ray imaging --potential advantages over

absorption:

• Differences between object and vacuum: small in both cases, but larger for n than for α

• This advantages increases as the wavelength decreases

• Better edge visibility, better contrast, smaller dose


Examples of “refraction” radiology:

X-rays images of a 0.5 mm live microfish


Building on bubbles (zinc electrodeposition):

QuickTime™ et undécompresseur H.264

sont requis pour visionner cette image.


Localized Electrochemical Deposition (LECD): a novel technique for growing high-aspect-ratio nanostructures:

Below a “critical” value of the microelectrode-structure distance, the growth rate

increases dramatically but the grown microstructure

becomes porous

This effect depends on the applied voltage


Cells in a leaf skin

Coherent x-ray micrographs of cells

Neurons from a mouse brain

Cultured fibroblast cells from a rabbit bone


Opening of a “stoma”




A bit more sophisticated description

Coherent* source

In the actual image, each edge is marked by fringes produced by Fresnel edge

diffraction. The fringes enhance the edge and

carry holographic information

Object

Detector

* Small &collimated


Modeling: interplay of “refraction” and “diffraction”

Refraction radiographs

Diffraction radiographs

Note: with bending-magnet emission, the effects are only in the vertical direction (no space coherence in the horizontal direction)


Coherent x-ray tomography:


Phase contrast microtomography:

microfossil

Li Chai Wei, Yeukuang Hwu, Jung Ho Je et al.


Phase contrast micro-tomography: rat kidney



Yeukuang Hwu, Jung Ho Je et al.


Phase contrast micro-tomography: rat aorta

Yeukuang Hwu, Jung Ho Je et al.




New types of sources:• Ultrabright storage rings (SLS, new

Grenoble project) approaching the diffraction limit

• Self-amplified spontaneous emission (SASE) X-ray free electron lasers

• VUV FEL’s (such as CLIO)• Energy-recovery machines • Inverse-Compton-scattering table-top

sources


Take a standard photon source with limited brightness and no

lateral coherence …… with a pinhole (size ξ), we can extract coherent light with good

geometrical characteristics (at the cost of losing most of the

emission)ξ

However, if the pinhole size is too small diffraction effects increase the beam

divergence so that:

ξθ >λθ

No source geometry beats this diffraction limit


Infraredphoton hν

Electron,energy = γ

γ’ ≈ γX-rayphoton hν’

Doppler effect: in the electron beam frame, the photon energy ≈ 2γ hν.This is also the energy of the backscattered photon in the electron-beam frame.

In the laboratory frame, there is again a Doppler shift with a 2γ factor, thus:

hν' ≈ 4γ2 hν


Energy-recovery LINAC sources

The brightness depends on the geometry of the source, i.e., of the electron beam

In a storage ring, the electrons continuously emit photons.

This “warms up” the electron beam and negatively affects its

geometry

Controlling the electron beam geometry is much easier in a linear accelerator (LINAC). Thus, LINAC sources can reach higher brightness levels


Energy-recovery LINAC sources

However, contrary to the electrons in a storage ring, the electrons in a LINAC produce photons only once: the power cost is too high

Solution: recovering energy

Accelerating section

Energy-recovery section


Example: Kulipanov’s “super-microtron” ER LINAC

Wiggler


THE “4 GLS” CONCEPT AT DARESBURY


Free-electron lasers:


Free-electron laser surgery:

Wavelength selection → much less collateral damage:


The scanning near-field opticalmicroscope (SNOM): like the stethoscope

Heart:Frequency ≈ 30-100 HzWavelength λ ≈ 102 m

Accuracy in localization ≈ 10 cm ≈ λ /1000

Small aperture

Small distance

Coated small-tip

optics fiber

Microscopic light-emitting object

SNOM resolution: well below the “diffraction limit” of standard microscopy (≈ λ)


∆y ∆ky > 2π → ∆y > 2π/∆ky

∆ky < ky = √(k2 - kx2)

kx real → ∆ky < k = 2π/λ

∆y > λ(diffraction limit)

SNOM: why does it work? Consider two slits:

x

y

Wave, k = 2π/λ

After a narrow optics fiber tip, there is an “evanescent wave” with imaginary in the

x-direction kx

However, for kx imaginary the condition does not apply and

∆y < λbecomes possible


20x20 µm2 SNOM image of growth medium (A. Cricenti et al.):

SNOM topography

S-O & N-Ovibrations

(λ = 6.95 µm)

λ = 6.6 µm

Intensity line scan

Resolution≈ 0.15 µm << λ


Self-amplified spontaneous emission x-ray free-electron lasers (SASE X-FEL’s)

Normal (visible, IR, UV) lasers:optical amplification in amplifying mediumplus optical cavity (two mirrors)

X-ray lasers: no mirrors → no optical cavity →need for one-pass high optical amplification

SASE strategy:

LINAC (linear accelerator)

Wigglerelectron bunch

The microbunching increases the electron density and the amplification and creates very short pulses


Seeding-Amplifier X-FELs

Electron BeamElectron Beam

Bypass

First Wiggler(SASE Emitter)

Second Wiggler (Amplifier)

Monochromator

Electron Dump

Photon Beam


First Real Experiments at the TESLA X-FEL’s


SASE X-FEL’s: superbright (orders of magnitude more than present sources) femtosecond pulses:

• New chemistry?• One-shot crystallography

(no crystals)?• Total coherence• Unprecedented

electromagnetic energy density

• Is this “vacuum”?


New physics?Consider the parameters of the Swiss Light Source:

Circumference: 288 m Single Bunch Current: 10-4 ABunch Length: 4 x 10-3 m Electron speed ≈ c ≈ 3.3 x 108 108 m/s

The charge per bunch is 10-4 x 288/(3 x 108) ≈ 10-8 coulomb, corresponding to 6 ∞ 1010 electrons.The horizontal bunch size is < 2 x 10-5. Assuming 0.1% coupling, the bunch volume is < 1.6 x 10-15 m3. Thus, the electron density exceeds 4 x 1019 cm-3.

What is this: a gas of independent electrons? Or a correlated multi-particle system?

What kind of thermodynamics should we use? The covalent form of thermodynamics is still an open issue!

For example: T can be defined using the entropy law or the equipartition principle. The two definitions are equivalent in classical physics, but in relativity they lead to different Lorentz transformations of T!


Conclusions:1. The technology of storage rings and FEL's solved the

ancient problem of brightness.2. The brightness increase was so rapid that applications

are still trailing behind.3. Nevertheless, many exciting results were obtained, for

example in spectromicroscopy and high resolution spectroscopy.

4. The most important new achievements will be linked to the interdisciplinary use of photon sources — exporting physics and chemistry techniques to medical research and the life sciences in general.

5. Coherence-based applications will play a special role.6. New FEL's like the SASE machines are beyond

imagination: towards one-shot crystallography?


Thanks:• The EPFL colleagues (Marco Grioni, Davor

Pavuna, Laszlo Forro Mike Abrecht, Amela Groso, Luca Perfetti, Eva Stefanekova, Slobodan Mitrovic, Dusan Vobornik, Helmuth Berger, Daniel Ariosa...).

• The POSTECH colleagues (group of Jung Ho Je).• The Academia Sinica Taiwan colleagues (group of

Yeukuang Hwu).• The Vanderbilt colleagues (group of Norman Tolk).• The ISM-Frascati colleagues (groups of Antonio

Cricenti and Paolo Perfetti)• The facilities: PAL-Korea, Elettra-Trieste.

Vanderbilt FEL, SRRC-Taiwan, APS-Argonne, SLS-Villigen, LURE-Orsay


• In 1905, Albert Einstein published his landmark articles about relativity, the photon (and the photoelectric effect) and the Brownian motion (demonstration of the existence of atoms and molecules)

• In our research, we use relativity to produce x-rays (photons) and use them, for example performing experiments with the photoelectric effect, to study solids, atoms and molecules

• HAPPY 2005!!!

A final remark:

Luce di sincrotrone: generazione e...

Documents

Transcript of Luce di sincrotrone: generazione e...